Lemon-Fruit-Based Green Synthesis of Zinc Oxide Nanoparticles and Titanium Dioxide Nanoparticles against Soft Rot Bacterial Pathogen Dickeya dadantii

Edible plant fruits are safe raw materials free of toxicants and rich in biomolecules for reducing metal ions and stabilizing nanoparticles. Zinc oxide nanoparticles (ZnONPs) and titanium dioxide nanoparticles (TiO2NPs) are the most produced consumer nanomaterials and have known antibacterial activities but have rarely been used against phytopathogenic bacteria. Here, we synthesized ZnONPs and TiO2NPs simply by mixing ZnO or TiO2 solution with a lemon fruit extract at room temperature and showed their antibacterial activities against Dickeya dadantii, which causes sweet potato stem and root rot disease occurring in major sweet potato planting areas in China. Ultraviolet–visible spectrometry and energy dispersive spectroscopy determined their physiochemical characteristics. Transmission electron microscopy, scanning electron microscopy, and X-ray diffraction spectroscopy revealed the nanoscale size and polymorphic crystalline structures of the ZnONPs and TiO2NPs. Fourier-transform infrared spectroscopy revealed their surface stabilization groups from the lemon fruit extract. In contrast to ZnO and TiO2, which had no antibacterial activity against D. dadantii, ZnONPs and TiO2NPs showed inhibitions on D. dadantii growth, swimming motility, biofilm formation, and maceration of sweet potato tuber slices. ZnONPs and TiO2NPs showed similar extents of antibacterial activities, which increased with the increase of nanoparticle concentrations, and inhibited about 60% of D. dadantii activities at the concentration of 50 µg∙mL−1. The green synthetic ZnONPs and TiO2NPs can be used to control the sweet potato soft rot disease by control of pathogen contamination of seed tubers.


Introduction
Nanoparticles (NPs) with at least one dimension in the range of 1-100 nm have high surface-to-volume ratios and display exceptional physical, chemical, and biological properties compared

Lemon Fruit Extract
Fresh lemon (Citrus limon) fruits bought from the supermarket were washed with tap water and Millipore water, cut into pieces, and dried at 60 • C for 10 h. Dry lemon pieces were ground into powder and mixed with Millipore water (1 g with 100 mL, Millipore, Molsheim, France), and then stirred continuously at 100 rpm at 60-70 • C for 4 h. After cooling to room temperature, the suspension was filtered through muslin cloth and then Whatman No. 1 filter paper; the extract (assumed as 10 mg·mL −1 ; pH about 4) was used for the synthesis of NPs or stored at −80 • C.

Synthesis of ZnONPs and TiO 2 NPs
ZnO and TiO 2 (analytical grade, purity ≥98%) (Sinopharm, Shanghai, China) were used for synthesis of ZnONPs and TiO 2 NPs, respectively. ZnO solution (0.5 M) and TiO 2 solution (0.5 M) were prepared by dissolving ZnO (4.07 g) and TiO 2 (4.00 g) separately in ethylene glycol (10 mL) (Sinopharm) and adding Millipore water to 100 mL. ZnONPs and TiO 2 NPs were separately synthesized using a protocol modified from a previous study [24]. The metal oxide solution (50 mL) was mixed with the extract of lemon fruits (50 mL) at the ratio 1:1 in flasks at 100 rpm at room temperature for 4 h and became colloid. After mixing the colloid (2 mL) with Millipore water (2 mL), the colloidal NPs were identified by ultraviolet-visible spectroscopy with a Shimadzu UV-2550 spectrometer (Shimadzu, Kyoto, Japan) from 200 to 800 nm at 1 nm resolution. The colloidal NPs were centrifuged at 27,200 g for 10 min and the pellets were washed with Millipore water and then freeze-dried with an Alpha 1-2 LDplus (Martin Christ GmbH, Osterode am Harz, Germany). The freeze-dried NPs were stored at −80 • C or prepared as stock solutions (50 mg·mL −1 ) for further analyses.

Characterization of ZnONPs and TiO 2 NPs
Dry lemon powder, ZnONP powder, and TiO 2 NP powder were analyzed by Fourier transform infrared (FTIR) spectroscopy to detect groups responsible for synthesis and stabilizing ZnONPs and TiO 2 NPs as previously described [41]. Bruker infrared table (Bruker Optics Inc. Billerica, MA, USA) and LibreTexts infrared spectroscopy absorption table (https://chem.libretexts.org/) were used to interpret the FTIR spectra. The elements of ZnONPs and TiO 2 NPs were detected by energy dispersive X-ray spectroscopy [41]. The size and morphology and the crystalline nature of ZnONPs and TiO 2 NPs were observed and analyzed by transmission electron microscopy (TEM), scanning electron microscopy, and X-ray diffraction spectroscopy [41].

Determination of Antibacterial Activities of Nanoparticles
Dickeya dadantii CZ1501 grown to mid-exponential phase was adjusted with the nutrient broth to about 5 × 10 8 CFU·mL −1 before use.
Antibacterial activities against D. dadantii were first detected by the diffusion assay with agar plates [41]. Dickeya dadantii suspension was inoculated into the nutrient agar to 10 7 cells·mL −1 . Wells (7 mm in diameter) were made in the agar plates with sterilized steel punchers. Fifty microliters of lemon fruit extract (10 mg·mL −1 ), ZnO (0.5 M), TiO 2 (0.5 M), ZnONP (50 µg·mL −1 ), and TiO 2 NP (50 µg·mL −1 ) were loaded into the wells and incubated at 30 • C for 24 h. Antibacterial activities were determined by the diameters of the clearing zones formed around the wells.
Dickeya dadantii suspension in nutrient broth (1 × 10 7 CFU·mL −1 ) without or with lemon fruit extract, ZnO, TiO 2 , ZnONPs, or TiO 2 NPs prepared as described above was transferred into wells (200 µL of suspension in each well) of 96-well microplates (Corning-Costar Crop., Corning, NY, USA) and incubated at 30 • C for 24 h. Biofilm formed by D. dadantii was stained by crystal violet and quantified by absorbance at 590 nm as previously described [41].
The swimming motility of D. dadantii was determined with semisolid nutrient agar (0.3% (w/v)) [41]. Lemon fruit extract, ZnO, or TiO 2 was added into the semisolid nutrient agar to the final concentration of 50 µg·mL −1 . ZnONPs or TiO 2 NPs was added into the semisolid nutrient agar to final concentrations of 12, 25, and 50 µg·mL −1 . Dickeya dadantii suspension (5 µL) was spotted onto the center of the semisolid nutrient agar plates and incubated at 30 • C for 48 h. The diameters of the halo-like colonies of D. dadantii were measured [41].

Statistical Analysis
All data in each experiment were subjected to one-way analysis of variance and means were compared by the Duncan's multiple range test using the SPSS software version 16 (SPSS, Chicago, IL, USA). The significance was set at p < 0.05.

Characterization of ZnONPs and TiO 2 NPs Synthesized with Lemon Fruit Extract
ZnONPs and TiO 2 NPs were synthesized by constant mixing of ZnO or TiO 2 solution with the lemon fruit extract at room temperature. ZnONPs displayed a characteristic surface plasmon resonance peak at 388 nm in the range of 350-420 nm (Figure 1a) determined by ultraviolet-visible spectroscopy [24]. TiO 2 NPs displayed a characteristic surface plasmon resonance peak at 410 nm in the range of 360-450 nm (Figure 1a) [23].
The FTIR spectrum of the ZnONPs (Figure 1b  , where D is the average particle size, K is the Scherrer constant (0.9), λ is the X-ray wavelength (0.15406 nm), β is the full width at half maximum of the X-ray diffraction peak, and θ is the Bragg angle.

Antibacterial Activity of ZnONPs and TiO 2 NPs Against Dickeya dadantii
Lemon fruit extract (10 mg·mL −1 ), ZnO (0.5 M), and TiO 2 (0.5 M), the raw materials for synthesis of ZnONPs and TiO 2 NPs, did not generate clearing zones around them in the nutrient agar containing D. dadantii cells and thus did not inhibit the growth of D. dadantii. In contrast, the products ZnONPs (50 µg·mL −1 ) and TiO 2 NPs (50 µg·mL −1 ) generated clearing zones with diameters of 30.0 ± 0.7 mm and 28.5 ± 0.5 mm, respectively, around the wells (including the well diameter 7 mm) and thus inhibited the growth of D. dadantii or killed D. dadantii. In nutrient broth, ZnONPs and TiO 2 NPs significantly inhibited D. dadantii growth and the extents of inhibition increased with the increase of the concentrations (12, 25, and 50 µg·mL −1 ) of ZnONPs and TiO 2 NPs, whereas lemon fruit extract, ZnO, or TiO 2 (50 µg·mL −1 ) did not inhibit D. dadantii growth (Figure 3a). ZnONPs and TiO 2 NPs inhibited D. dadantii growth at similar extents.
Dickeya dadantii grew and swam in the semisolid nutrient medium and formed a halo-like colony about 24 mm in diameter after 48 h (Figure 3b). Lemon fruit extract, ZnO, or TiO 2 (50 µg·mL −1 ) did not inhibit D. dadantii growth and swimming in the semisolid medium, whereas ZnONPs and TiO 2 NPs significantly inhibited D. dadantii growth and swimming and the extents of inhibition increased with the increase of the concentrations of ZnONPs and TiO 2 NPs (Figure 3b). ZnONPs and TiO 2 NPs inhibited D. dadantii growth and swimming motility at similar extents, about 31-32%, 44%, and 60% at the concentrations of 12, 25, and 50 µg·mL −1 , respectively.
Dickeya dadantii cells formed biofilms on the surface of the polystyrene microplate wells during the 24 h incubation. Lemon fruit extract, ZnO, or TiO 2 (50 µg·mL −1 ) did not inhibit the biofilm formation, whereas ZnONPs and TiO 2 NPs significantly inhibited the biofilm formation and the extents of inhibition increased with the increase of the concentrations of ZnONPs and TiO 2 NPs (Figure 3c). ZnONPs and TiO 2 NPs inhibited D. dadantii biofilm formation at similar extents, about 34-37%, 54-55%, and 64-66% at the concentrations of 12, 25, and 50 µg·mL −1 , respectively.   Transmission electron microscopy revealed the morphological changes of D. dadantii cells after growing in the nutrient broth with ZnONPs and TiO2NPs (50 µg•mL -1 ). -Dickeya dadantii cells grown without NPs had intact cell envelopes and dense cytoplasm filled in the cells (Figure 4,a-b). After 4 h growth with NPs, cell envelopes of many D. dadantii cells became distorted and disintegrated while cytoplasm became shrunken, agglomerated, and collapsed (Figure 4,c-f), leading to cell death.

Discussion
We made ZnONPs and TiO2NPs separately by mixing ZnO or TiO2 dissolved in ethylene glycol and water with the lemon fruit extract at room temperature. Energy dispersive X-ray spectroscopy on the NPs revealed C element from the lemon fruit extract. Fourier transform infrared spectroscopy revealed that the functional groups on the surface of the ZnONPs and TiO2NPs and responsible for stabilization of the ZnONPs and TiO2NPs were from the lemon fruit extract. The major O-H and N-H groups are associated with carboxylic acids such as citric acid and ascorbic acid [13], and amines such as free amino acids and proteins, which are rich in lemon fruits [8]. The O-H groups of alcohol may be from polyols in the lemon fruit extract and the solvent ethylene glycol [44]. Citric acid has three carboxylate groups and is able to form stable complexes with metal ions. Citric acid has been used as a reducing and stabilizing agent in synthesis of a wide range of nanomaterials to control both the size and morphology of the nanomaterials [13,45,46]. Ethylene glycol with two hydroxyl groups has a relatively strong reducing powder and high boiling point, and has been widely used in polyol synthesis of metal nanomaterials [44,47]. Unlike TiO2, ZnO is not stable in acidic solutions [48].

Discussion
We made ZnONPs and TiO 2 NPs separately by mixing ZnO or TiO 2 dissolved in ethylene glycol and water with the lemon fruit extract at room temperature. Energy dispersive X-ray spectroscopy on the NPs revealed C element from the lemon fruit extract. Fourier transform infrared spectroscopy revealed that the functional groups on the surface of the ZnONPs and TiO 2 NPs and responsible for stabilization of the ZnONPs and TiO 2 NPs were from the lemon fruit extract. The major O-H and N-H groups are associated with carboxylic acids such as citric acid and ascorbic acid [13], and amines such as free amino acids and proteins, which are rich in lemon fruits [8]. The O-H groups of alcohol may be from polyols in the lemon fruit extract and the solvent ethylene glycol [44]. Citric acid has three carboxylate groups and is able to form stable complexes with metal ions. Citric acid has been used as a reducing and stabilizing agent in synthesis of a wide range of nanomaterials to control both the size and morphology of the nanomaterials [13,45,46]. Ethylene glycol with two hydroxyl groups has a relatively strong reducing powder and high boiling point, and has been widely used in polyol synthesis of metal nanomaterials [44,47]. Unlike TiO 2 , ZnO is not stable in acidic solutions [48]. Dissolution of ZnO to Zn 2+ may occur after mixing ZnO solution with the lemon fruit extract (pH 4.0) while Zn 2+ may be chelated by citrate through two carboxyl groups and one hydroxyl group and form a pentabasic ring and a hexahydric ring [49]. Perhaps, esterification of citric acid and ethylene glycol and binding between esters and Zn 2+ may occur in the mixture and stabilize Zn 2+ [16]. Citric acid may also reduce the surface tension of ZnO and TiO 2 solutions and lower the energy needed to form the ZnO and TiO 2 crystals [49]. Together, multiple carboxylic acids, amino acids, and polyols may lead to the formation of the polymorphic ZnONPs and TiO 2 NPs.
In contrast to ZnO and TiO 2 with no antibacterial activity against D. dadantii, ZnONPs and TiO 2 NPs showed distinct antibacterial activities against D. dadantii growth, swimming motility, biofilm formation, and maceration of sweet potato tuber slices. ZnONPs and TiO 2 NPs showed similar extents of antibacterial activities against D. dadantii and an increase of antibacterial activities with the increase of NP concentrations, and inhibited about 60% of D. dadantii growth, swimming motility, biofilm formation, and maceration of tuber slices at the concentration of 50 µg·mL −1 .
The distinct antibacterial activities of metal and metal oxide NPs were achieved by their smaller sizes and larger surface-area-to-mass ratios and generation of oxidative stress on bacterial cells [21,50]. The smaller size and larger surface area lead NPs to easily adsorb bacterial cells and a higher proportion of atoms on the particle surface, and enhance the ability to pass through membranes and the interfacial reactivity to directly interfere with cell envelope functions. Metal oxides like ZnO and TiO 2 have known photo-oxidizing and photocatalytic activities and generate reactive oxygen species (ROS). ZnONPs and TiO 2 NPs adhering to the cell surface can generate extracellular ROS and induce intracellular generation of ROS, leading to oxidative stress on cells, distortion and damage of cell membranes, leakage of intracellular contents, and eventually cell death [21,50]. In this study, in vitro and in vivo actions of ZnO and TiO 2 and their NPs on D. dadantii were processed in the dark; photo-oxidization and photocatalysis may not participate in the antibacterial mechanisms of ZnONPs and TiO 2 NPs against D. dadantii. ZnONPs and TiO 2 NPs may induce osmotic stress on bacterial cells in the dark and induce membrane depolarization and loss of membrane integrity, resulting in cellular leakage [51], as revealed by TEM. In addition, Zn 2+ may be dissolved from ZnONPs and expose its heavy metal toxicity to cells.
The ZnONPs and TiO 2 NPs produced in this study show polymorphic structures and surface defects with numerous edges and pits. Such abrasive surface texture not only has more reactive surface sites but also tends to disrupt cell membranes and abrade biofilms. D. dadantii regulates plant surface colonization via regulation of flagella-mediated motility and biofilm formation. D. dadantii movement and attachment to plant surfaces and formation of biofilms on plant surfaces, in intercellular spaces, and in xylem vessels are essential for survival and completing disease cycles [52][53][54][55]. Perhaps, ZnONPs and TiO 2 NPs adhering to bacterial cell surfaces may inhibit bacterial movement and attachment to plant surfaces, inhibit bacterial growth, and damage bacterial cells and biofilms.
We previously produced AgNPs against D. dadantii by mixing AgNO 3 solution with cell-free bacterial culture supernatants for 48 h [41]. That entire process required screening of bacterial culture supernatants and sterile operation to avoid contamination, which was time-consuming and not laborand cost-effective as compared to this green-synthesis with the lemon fruit extract.
The reason we used lemon fruit extract to produce ZnONPs and TiO 2 NPs instead of AgNPs is to avoid using toxic Ag + . However, the green synthetic ZnONPs and TiO 2 NPs have properties distinct from ZnO and TiO 2 , such as the antibacterial activity against D. dadantii, and raise the concern about NP toxicity to plants and humans. NPs can be taken up and accumulated in plants and enter the food chains of animals and human, and thus pose a risk to human health [20,56,57].
ZnONPs and TiO 2 NPs may have positive and negative impacts on plants, depending on not only NP properties (size, shape, surface coating, and stability), concentration, and exposure time but also plant properties (susceptible or tolerant to NPs) and development stages. Generally, ZnONPs and TiO 2 NPs in excess are harmful to plants, while in traces can be beneficial for plants [56][57][58]. Bradfield et al. [59] grew sweet potato to maturity in field microcosms using substrate amended with either ZnONPs, CuONPs, or CeO 2 NPs or equivalent amounts of Zn 2+ , Cu 2+ , or Ce 4+ at three concentrations (100, 500, or 1000 mg·kg DW −1 ). Only the application with the highest concentration of Zn or Cu, which is unlikely to occur in the environment, caused adverse outcomes, that is, reduction in tuber biomass, metal accumulation, and dietary intake regardless of the chemical forms of the metals added to the substrate. The concentrations of metals were higher in the peels than in the inner tuber flesh of sweet potato. Under such conditions, metal oxide NPs pose no greater risk to sweet potato yield or food safety than do the ionic metals [59]. Bonilla-Bird et al. [60] immersed sweet potato tubers in suspensions or solutions of CuONPs, CuO, and CuCl 2 at 25, 75, or 125 mg·L −1 under continuous stirring for 30 min and found that the Cu concentration in internal tissues of tubers treated with CuONPs was similar to that in control tubers, suggesting no risk of Cu contamination in peeled tubers. The peel of sweet potato tubers restricts the inward radial transfer of metal NPs. Thorough cleaning of tuber surfaces and peeling of tubers can effectively reduce the consumer exposure to metal NPs or ionic metals [59].

Conclusions
We used ZnO and TiO 2 , which are generally recognized as safe substances, and extracts from edible lemon fruits, which are free of toxicants and rich in biomolecules, as safe raw materials to reduce health and environmental risks of the sources for the production of NPs. We developed a simple, rapid, cost-effective, and ecofriendly method to produce green ZnONPs and TiO 2 NPs simply by mixing ZnO or TiO 2 solution with the lemon fruit extract. Actions of carboxylic acids, amino acids, and polyols in the lemon fruit extract may lead to the formation of the polymorphic ZnONPs and TiO 2 NPs. In contrast to ZnO and TiO 2 , the ZnONP and TiO 2 NP products effectively inhibited D. dadantii growth, swimming motility, biofilm formation, and maceration of sweet potato tubers and likely combat the bacterial pathogen via multiple mechanisms. The multiple bacteriocidal and bacteriostatic mechanisms would make it difficult for the pathogen to develop resistance to the NPs. The latently infected seed tuber is one of the major sources of the soft rot disease of sweet potato. The green ZnONPs and TiO 2 NPs appear to be promising materials to treat the seed tubers to avoid and reduce the pathogen contamination and to produce healthy crops. A study is needed to clarify if ZnONPs and TiO 2 NPs may contaminate the sweet potato tuber flesh and can be used to preserve and increase the shelf life of sweet potato tubers without exposing consumers to excess metals. More studies are needed before bringing the ZnONPs and TiO 2 NPs to the field to control the soft rot disease and promote sweet potato growth.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.